Hi Tomás,

You wrote:

> After some research we found out that we were not rotating those
> logs so we added to the zoo.cfg the following variables:
>
> autopurge.snapRetainCount=3
> autopurge.purgeInterval=1

The documentation is as follows:

>> *autopurge.snapRetainCount* : [...] WhenZooKeeper auto purge feature
>> retains the **autopurge.snapRetainCount** most recent snapshots and
>> the corresponding transaction logs in the **dataDir** and
>> **dataLogDir** respectively and deletes the rest.

You are showing us the /transaction logs/ directory, which indeed
contains more than three files:

> $ cd /var/lib/zookeeper/data/log/version-2
> $ ls -la

but that may be perfectly normal if these are "held" by up to three
snapshots.  We would have to see that files are indeed accumulating in
the /snapshot/ directory.

HTH, -D



-----

Tomás Almeida <[email protected]> writes:
> Hello everyone,
> A few weeks ago our team detected that zookeeper had a volume that was
> consuming a lot of disk. We found out that the folder that was using too
> much disk was the dataDir where the data logs and the snapshots are being
> stored. After some research we found out that we were not rotating those
> logs so we added to the zoo.cfg the following variables:
>
> autopurge.snapRetainCount=3
> autopurge.purgeInterval=1
>
> What we discovered from the logs was strange:
> -The purge was running (every hour).
> -Eventhough it was running, it wasn't doing anything and the folder would
> get bigger and bigger.
>
> Environment:
> Running 3 zookeeper pods in OpenShift using a PVC to store this type of
> logs (the data logs and the snapshots).
> Zookeeper version 3.8.4
>
>
> Log snippet of purge running:
> 2024-12-12 12:14:12,224 [myid:] - INFO
> [PurgeTask:o.a.z.s.DatadirCleanupManager$PurgeTask@139] - Purge task
> started.
> 2024-12-12 12:14:12,224 [myid:1] - INFO [main:o.a.z.s.q.QuorumPeerMain@152]
> - Starting quorum peer, myid=1
> 2024-12-12 12:14:12,226 [myid:] - INFO
> [PurgeTask:o.a.z.s.p.FileTxnSnapLog@124] - zookeeper.snapshot.trust.empty :
> false
> 2024-12-12 12:14:12,324 [myid:] - INFO [PurgeTask:o.a.z.s.p.SnapStream@61]
> - zookeeper.snapshot.compression.method = CHECKED
> 2024-12-12 12:14:12,327 [myid:] - INFO
> [PurgeTask:o.a.z.s.DatadirCleanupManager$PurgeTask@145] - Purge task
> completed.
>
>
> Log folder:
> $ cd /var/lib/zookeeper/data/log/version-2
> $ ls -la
> total 608
> drwxrwsr-x. 2      9994 zookeeper    12288 Dec 12 12:15 .
> drwxrwsr-x. 3      9994 zookeeper     4096 Apr  9  2024 ..
> -rw-rw-r--. 1 zookeeper zookeeper 67108880 Dec  5 17:58 log.1e900000001
> -rw-rw-r--. 1 zookeeper zookeeper 67108880 Dec  6 08:23 log.1ea00000001
> -rw-rw-r--. 1 zookeeper zookeeper 67108880 Dec  7 15:59 log.1eb00000001
> -rw-rw-r--. 1 zookeeper zookeeper 67108880 Dec  9 10:23 log.1ec00000001
> -rw-rw-r--. 1 zookeeper zookeeper 67108880 Dec 10 18:34 log.1ed00000001
> -rw-r--r--. 1 zookeeper zookeeper 67108880 Dec 12 12:15 log.1ee00000001
>
>
> Can someone help me identify what is happening?
>
> Thank you all in advance!

Reply via email to